l-4/t-2/eee date: 09/01/2021 - lib.buet.ac.bd:8080

L-4/T-2/EEE Date: 09/01/2021

BANGLADESH UNIVERSITY OF ENGINEERING AND TECHNOLOGY, DHAKA

L-4/T-II B.Sc. Engineering Examination 2018-19

Sub: EEE 371 (Power System II)

Full Marks: 180 Time 2 Hours

The Figures in the margin indicate full marks

USE SEPARATE SCRIPTS FOR EACH SECTION

There are 3 page(s) in this question paper.

SECTION – A

There are FOUR questions in this section. Answer any THREE

All the symbols have their usual meanings

Assume reasonable values for missing data.

1. (a) Show that the transmission loss equation of a power system having K

number of generators can be expressed by the following equation.

K

i

K

j

K

i

giigjijgiL BPBPBPP1 1 1

000

(20)

(b) The incremental fuel costs (in $/MWhr) for a plant consisting of two units

are given by: λ1 = (0.008Pg1+8) and λ2 = (0.0096Pg2+6.40). The maximum and

minimum generation capacities of both units are 625 MW and 100 MW

respectively. Determine the annual saving in fuel cost (in $) for the economic

distribution of a total load of 900 MW between units compared with the

following distribution of the load: unit-1: 500 MW and unit-2: 400 MW.

(10)

2. (a) Show that the Fast Voltage Stability Index (FVSI) of k-th bus can be

expressed by the following relation.

XV

QZFVSI

S

kk 2

24

(15)

(b) (i) What is voltage sag? How is it calculated due to a fault?

(ii) Explain the design principle of a single tuned passive filter for harmonics

mitigation.

(15)

= 2 =

3. (a) Define the commonly used performance metrics of frequency response.

Explain the role of governor response and under frequency load shedding in the

context of frequency response adequacy.

(15)

(b) How does load frequency relief (kp) contribute to frequency stability?

Formulate a general expression of kp from the following exponential load

model.

0

1

0

0 1f

Δfk

V

VPP p

np

(15)

4. (a) Briefly explain the working principle of a STATCOM and a UPFC. (15)

(b) A ± 30 MVAR SVC is delta connected to 132 kV system via a 132/6 kV

transformer. The SVC consists of a 12-pulse TCR and two switched capacitor

arms, one tuned to the 11th harmonic and the other tuned to the 13th harmonic

(the lowest harmonics produced by the TCR) and producing equal amounts of

reactive power. Calculate the values of the passive components required for the

TCR and the switched capacitor branches.

(15)

SECTION – B




5. (a) Define power system stability. Show the classification of power system

stability using a tree diagram.

(10)

(b) Derive the general expressions of power and torque of a salient pole

machine considering two-axis model.

(20)

6. (a) Derive the following relation which governs the rotational dynamics of a

synchronous machine in power system stability.

2 per unit

m e

s

H dP P

dt

(15)

(b) With necessary diagrams, explain the equal-area criterion for a typical

single machine infinite bus system when a fault occurs at the sending end bus.

(15)

= 3 =

7. (a) Derive a linear 2nd

order differential equation governing the incremental

variations in rotor angle and output power of a synchronous machine when the

mechanical input power is fixed.

(15)

(b) The single-line diagram of figure given below shows a generator connected

through parallel transmission lines to an infinite bus. The machine is delivering

1.0 per unit power and both the terminal and infinite-bus voltages are 1.0 per

unit. Numbers on the diagram indicate the values of reactances on a common

system base. The transient reactance of the generator is 0.2 per unit as

indicated. Calculate the value of synchronizing power coefficient when the

machine is operating at 28.44o and is subjected to a slight temporary

electrical system disturbance

Figure for Q.7(b)

(15)

8. (a) Explain the factors which affect the transient stability of a power system. (10)

(b) Describe the step-by-step solution procedures of swing equations for a

multi-machine power system.

(20)

L-4/T-2/EEE Date: 19/01/2021


L-4/T-2 B.Sc. Engineering Examination 2018-19

Sub: EEE 425 (Biomedical Instrumentation)




All the symbols have their usual meanings. Assume reasonable values for missing data.


SECTION – A


1. (a) Sketch the action potentials corresponding to five major portions of the

heart during a cardiac cycle and also sketch the resulting ECG generated from

those five action potentials (maintain a relative time scale).

(18)

(b) Find the heart rate, PR interval, and QRS duration in the first ECG

waveshapes shown below in Fig 1 (b). State the possible type of disease and

mention etiology.

Fig 1 (b)

(12)

= 2 =

2. (a) From the following two ECG lead plots in Fig 2 (a) (i) and Fig 2 (a) (ii),

find whether the person corresponding to each plot is having a myocardial

infraction. If so, where?

Fig 2 (a) (i)

Fig 2 (a) (ii)

(18)

(b) From the following ECG lead plot in Fig 2 (b), mention the name of the

change seen and comment on what change is manifested compared to the

(12)

= 3 =

normal rhythm, under which leads and which location or area of the heart.

Fig 2 (b)

3. (a) Define defibrillator. Compare different kind of defibrillators. (10)

(b) From the following Fig 3 (b) of Sphygmomanometer cuff, mention the

relation between cuff pressure and blood flow in each stage.

Fig 3 (b)

(20)

4 (a) Explain the difference between ERP and EGG? Compare between

isomorphic and metamorphic seizure patterns.

10

(b) Draw the block diagram of the atrial synchronization pacemaker. While

explaining the working principle of atrial synchronization pacemaker, mention

its difference from demand mode pacemaker.

20

= 4 =

SECTION – B


5. (a) The recorded ECG signal shown below is affected by noise. Identify the

type of noise. In general filtering may reduce the effect of this noise. Briefly

discuss the effect of different types of filtering with possible outcome.

Fig. for Q. 5(a)

(15)

(b) With neat sketch, compare the functions of colorimeter, flame photometer

and spectrophotometer.

(15)

6. (a) Compare, with equivalent circuits, resistive, inductive, capacitive and

piezoelectric transducers.

(15)

(b) Draw the equivalent circuit of the body surface electrode. Compare the

working principles of different body surface recording electrodes.

(15)

7. (a) With sketch, mention why a collimator is installed in an X-ray machine?

Mention the interpretation of strong blackening and weak blackening in X-

ray image. These are manifested in which issues?

(15)

(b) Compare the principles of X-ray, CT and Ultrasound imaging. Write

interpretation of different scanning modes of ultrasound imaging.

.

(15)

8. (a) Mention applications of scintigraphy in lung, renal, brain and thyroid

problem diagnosis.

(15)

(b) Compare static and dynamic imaging techniques (SPECT, PET). Show with

diagram the function of the gamma camera (from the patient to the image).

(15)

L-4/T-2/EEE Date: 09/01/2021



Sub: EEE 437 (Wireless Communication)


The Figures in the right margin indicate full marks


There are 2 (TWO) pages in this question paper.

SECTION – A

There are FOUR questions in this section. Answer any THREE.

All the symbols have their usual meanings.


1. (a) Explain the two criteria used for ST code design. (15)

(b) Summarize the advantages and disadvantages of OSTBCs. Explain one OSTBC

and show why it is called an orthogonal code.

(15)

2. (a) What is the desired auto-correlation property of a PN code? Compare m-sequence

and Gold sequence in terms of auto-correlation property.

(12)

(b) Consider a 3-stage m-sequence generator having a characteristic polynomial (1 +

x + x2). If the initial seed is 001, determine the m-sequence. Also, write the m-

sequences if 010, 100, 110 and 111 are used as seeds.

(18)

3. (a) Explain the operation of an MC-DS-CDMA transmitter. (10)

(b) Find the channel capacity for the channel having the following transition matrix:

(20)

4 (a) Write short notes on - (i) LTE frame structure, and (ii) transport block formation

in LTE data/user plane.

(16)

(b) Consider a flat-fading channel with i.i.d. channel gain gi, which can take on three

possible values: g1 = 0.05 with probability p1 = 0.1, g2 = 0.5 with probability p2 =

0.5, and g3 = 1 with probability p3 = 0.4. The transmit power is 10 mW, the noise

power spectral density is N0 = 10−9 W/Hz, and the channel bandwidth is 30 KHz.

Assume that both the transmitter and the receiver have the knowledge of the

instantaneous value of gi. Find the ergodic capacity of this channel considering

adaptive power allocation using Water-filling algorithm.

(14)

= 2 =

SECTION – B


All the symbols have their usual meanings.

5. (a) Discuss the factors that affect the small-scale fading and large-scale fading.

Also, briefly mention the techniques used to overcome the effects of large-scale

fading and small-scale fading on the performance of wireless communication.

(15)

(b) A base station is transmitting 20W using a carrier frequency of 2.1 GHz

having bandwidth of 180 kHz. Antenna heights at the base station and the mobile

station are 25m and 2.5m, respectively. Antenna gains of the base station and

mobile station are 8 dB and 0 dB, respectively. The mobile station requires a

minimum SNR of 30 dB for proper communication. Calculate the (i) cell radius,

(ii) coverage area, and (iii) the data rate at a mobile station located at the cell

edge, assuming 2-ray ground reflection path-loss model.

(15)

6. (a) Define the average probability of symbol error for a fading channel. Then

derive the expression of average bit error rate for BPSK modulation considering

a Rayleigh fading channel.

(15)

(b) Consider a communication system with M diversity branches, where

instantaneous SNR of each diversity branch is exponentially distributed with

parameter 0.04. Determine which diversity combining technique out of SC, MRC

and EGC performs the best in this channel.

(15)

7. (a) Distinguish the signal models of MISO channel and MIMO channel. (10)

(b) When does the Water-filling algorithm become useful in MIMO

communication system? Explain the operation of the algorithm in such scenario.

(20)

8. (a) Explain the difference between the coding gain and diversity gain. (10)

(b) Prove that without using pre-processing, transmit antenna diversity (MISO)

does not provide any diversity gain. Then show that Alamouti scheme based pre-

processing scheme in MISO system provides full diversity gain. Assume that the

channel is unknown to the transmitter.

(20)

== 1 ==

L-4/T-2/EEE Date: 12/01/2021



Sub: EEE 443 (Radar and Satellite Communications)


The Figures in the right margin indicate full marks.


There are 4 pages in this question paper.

SECTION – A


Answer in brief and to the point. All the symbols have their usual meanings.

1. (a) What is a satellite? Define a satellite orbit and its inclination angle. Compare

and contrast the uplink and downlink frequencies of satellite communication

with mobile cellular communication.

Draw the block diagram of an FDM-TDM/PCM converter used in a digital

earth station to serve the terrestrial analog FDM customers and explain its

operation in brief.

(7+8)

(b) Mention the two unique features of satellite communication. Draw the block

diagram of a typical satellite repeater/transponder and mention the basic

functions of each block.

An earth station, operating at 12 GHz with a 10° elevation angle has a 47-dB

gain and 2.5 dB loss from the antenna feed to the input of the LNA. The sky

noise is 25° K developing an antenna noise temperature of 240° K. The noise

figure of the LNA is 1.5 dB. Calculate the G/T.

(7+8)

2. (a) With neat sketches, name and describe the use of the 4 types of satellite

antenna. Mention the special advantage of a phased-array antenna.

Draw the simplified diagram of the communication system carried by a typical

Intelsat satellite and mention the function of the switch matrix in it.

(8+7)

(b) Name the two types of earth-station antenna. With a neat sketch, describe

the functions of each part of any one of them.

What is a VSAT? With a neat sketch, describe its basic operation in brief.

Enumerate the applications of mobile satellite communications.

(9+6)

== 2 ==

3. (a) Explain the terms “FDM-FM-FDMA”, “TDM-PSK-TDMA” and

“TDM-SCPC-FDMA” with reference to satellite communication.

How can we increase the transponder bandwidth using SDMA/frequency reuse?

(9+6)

(b) A LEO satellite communication system uses direct sequence CDMA as the

multiple access method for groups of terminals within each of its multiple

antenna beams. The terminals generate and receive compressed digital voice

signals with a bit rate of 9.6 kbps. The signals are transmitted and received at a

chip rate of 5.0Mbps as BPSK modulated DSSS-CDMA. In the absence of any

other CDMA signals, the input power level at the receiver input is −116.0dBm

(−146.0dBW) for one CDMA signal, and the noise temperature of the receiving

system is 300°K. The satellite transmits 31 simultaneous CDMA signals.

(i) Find the SNR for the 9.6 kbps BPSK signal after despreading, and estimate

the BER of the data signal, given a system implementation margin of 1dB.

(ii) If two of the multiple beams from the satellite overlap, so that a second

group of 31 DS-SS CDMA signals is present at the receiver, find the BER

of the desired signal.

(8+7)

4. (a) Show a simplified diagram of a generic TDMA frame and mention how its

contents are used for synchronization between the earth stations and satellites in

a network.

With neat diagrams, show the star and mesh configurations of satellite

networks, and mention the basic functional differences between the two.

(8+7)

(b) Enumerate the satellite communication link design steps. Write down the

link equation from an earth station to a satellite and show its components in a

neat diagram.

Using the uplink noise power budget given in Table 1, first calculate the

required power at the transponder input, Pr (dBW), to meet the requirement of

(C/N)up=30dB. Finally, referring to the uplink power budget shown in Table 2,

calculate the uplink (earth station) transmitter power both in dB and watt. Is it a

realistic transmitter power? If not, suggest how it can be practically realized.

Table 1: Noise Power Budget (Uplink)

------------------------------------------------------------------------------------------------

Boltzmann’s constant , 1.38 x10–23

J/°K k –228.6 dBW/K/Hz

Transponder noise Temperature, 500°K TS 27.0 dBK

Receiver noise bandwidth, 43.2MHz B=Bn 76.4 dBHz

------------------------------------------------------------------------------------------------

Noise power in the transponder N=Pn –125.2 dW

(8+7)

== 3 ==

Table 2: Uplink (Carrier) Power Budget

------------------------------------------------------------------------------------------------

Earth station transmitter power Pt Pt dBW

Earth station antenna gain Gt 55.72 dB

Satellite receiving antenna gain Gr 31.0 dB

Free space path loss Lp –207.17 dB

Earth station on –2dB contour Lant –2.0 dB

Atmospheric path loss Lup –0.7 dB

Miscellaneous path loss Lmisc –0.3 dB

------------------------------------------------------------------------------------------------

Received power at transponder input Pr Pt –123.45 dBW

SECTION – B


All the symbols have their usual meanings. Assume reasonable values for missing data.

5. (a) What does a GPS transmitter transmit? How does it identify itself? (08)

(b) Explain the method of detecting position on or above the earth with the help

of a GPS Receiver. Include the necessary calculations for ideal conditions.

What are the major sources of errors in position calculations and the remedial

measures?

(22)

6. (a) Elaborate the term RADAR. What do you understand by range of a

RADAR? State five principal applications of RADAR.

(10)

(b) Answer the following questions:

i) State the frequency band commonly used for fire control and imaging.

ii) What is continuous wave RADAR?

iii) What is the most common waveform transmitted from a RADAR

transmitter?

iv) Distinguish between mono- and bi-static RADAR.

v) Write down the equation to measure RADAR range using the time

taken by the pulse to travel to the target and return. What is the

maximum unambiguous range?

(10)

(c) State intuitively with the help of diagram(s), why coherent integration

strengthens the backscattered signal but suppresses noise.

What is the role of a matched filter in achieving high range-resolution in

RADARs?

(10)

== 4 ==

7. (a) With the help of a block diagram, briefly describe each major component of

a modern RADAR system.

(15)

(b) Answer the following questions related to RADAR:

i) Other than doppler frequency shift method, name one method that can

be employed to determine the relative velocity of the target.

ii) In order to determine the shape of the target, what technique can be

employed? Name the type of RADAR that performs shape

reconstruction.

iii) Radio waves of HF band are refracted/reflected back to ground by

the ionosphere – mention one RADAR application of this phenomenon.

(09)

(c) For a RADAR system, the scattering matrix of the backscatterer is,

S =

.

Comment on the RADAR system on why it is single-polarized or fully

polarized.

(06)

8. (a) In SAR RADARs, how does a single antenna synthesize a large aperture?

Explain briefly.

(10)

(b) What is measured by the parameter RCS?

(c) A directive antenna with a gain of 6dB (with respect to an isotropic radiator)

at a particular direction transmits a power of 10 Watts at the operating

frequency of 1 GHz. Calculate the power density at a point 1 km away from the

antenna in the same direction. Now, if an airplane with an RCS value of 2 m2

appears at that point, calculate the power of the reflected signal at that position.

Also, calculate the power density of the reflected signal at the position of the

RADAR. Assume monostatic scenario.

(05)

(15)

-------------- 0 -------------

L-4/T-2/EEE Date: 23/01/2021



Sub: EEE 447 (Introduction to Digital Image Processing)


The figures in the margin indicate full marks. All the symbols have their usual meanings.




SECTION – A


1. (a) The white bars in the test pattern shown are 5 pixels in width and 80 pixels

in height. The separation between the bars is 10 pixels. Sketch the look of the

image after the application of a 55 median filter.

Fig. for Q.1(a)

(7)

(b) Let an 8-bit image 𝑓(𝑥, 𝑦) = (𝑥, 𝑦) be blurred by uniform motion 𝑥(𝑡) =

𝑎𝑡

𝑇 and 𝑦(𝑡) =

𝑏𝑡

𝑇, where a and b are non-negative scalars, and T is the time of

camera exposure. Design a Wiener filter for deblurring the degraded image.

Assume that white noise of variance 10 is added during the imaging condition.

(23)

2. (a) Sketch the Radon transform of the following square image. Assume a

parallel-beam geometry.

Fig. for Q. 2(a)

(7)

(b) Show that the Radon transform of the following Gaussian shaped 2D (23)

= 2 =

function

𝑓(𝑥, 𝑦) = (𝑥2 + 𝑦2)exp(−𝑥2 − 𝑦2)

is given by

𝑔(𝜌, 𝜃) =√𝜋

2(2𝜌2 + 1)exp(−𝜌2)

3. (a) Compute the Haar wavelet transform of a 2×2 image given by

𝑭 = [105 3533 107

]

(15)

(b) The inverse Haar transform is 𝑭 = 𝑯𝑻𝑾𝑯, where 𝑾 is the Haar wavelet

transform of the image and 𝑯𝑻 is the matrix inverse of 𝑯. Show that the

relation 𝑯𝟐−𝟏 = 𝑯𝟐

𝑻 exists and use it for computation of the inverse Haar

wavelet transform of the result found in (a)

(15)

4. Consider the simple 4 × 8, 8-bit image:

(i) Compute the entropy of the image considering the row-wise pairs of pixels.

(ii) Compress the image using the Huffman coding by considering the pairs of

pixels. Explain the result of compression.

(iii) Compress the image using LZW coding. Show details of the code words

and dictionary locations.

(iv) Compare the Huffman and LZW coding schemes with the results obtained.

(30)

SECTION – B


5. (a) What are the basic components of a general-purpose image processing

system? Briefly describe each component with necessary block diagram.

(20)

21 21 21 95 169 243 243 243 21 21 21 95 169 243 243 243 21 21 21 95 169 243 243 243 21 21 21 95 169 243 243 243

= 3 =

(b) A common measure of transmission for digital data is the baud rate, defined

as the number of bits transmitted per second. Generally, transmission is

accomplished in packets consisting of a start bit, a byte (8 bits) of information,

and a stop bit. Using these facts, answer the following: (i) How many minutes

would it take to transmit a 1024 × 1024 image with 256 intensity levels using a

56K baud modem? (ii) What would the time be at 3000K baud, a representative

medium speed of a phone DSL (Digital Subscriber Line) connection?

(10)

6. (a) Prove that both the 2-D continuous and discrete Fourier transforms are

linear operations.

(20)

(b) A camera is equipped with a 35 mm lens. The CCD camera chip of

dimensions 7 × 7 mm and having 1024 × 1024 elements, is focused on a square,

flat area, located 0.5 m away. How many line pairs per mm will this camera be

able to resolve?

(10)

7. (a) Show that 2-D transforms with separable, symmetric kernels can be

computed by (i) computing 1-D transforms along the individual rows (columns)

of the input, followed by (ii) computing 1-D transforms along the columns

(rows) of the result from step (i).

(15)

(b) An image with intensities in the range [0, 1] has the PDF pr(r) shown in Fig.

for 7(b). It is desired to transform the intensity levels of this image so that they

will have the specified pz(z) as shown. Assume continuous quantities and find

the transformation (in terms of r and z) that will accomplish this.

Fig. for Q. 7(b)

(15)

= 4 =

8. (a) Consider a 3 × 3 spatial mask that averages the four closest neighbors of a

point (x,y), but excludes the point itself from the average. Find the equivalent

filter H(u,v) in the frequency domain and show that H(u,v) is a low pass filter.

(15)

(b) The two images in Fig. for 8(b) are quite different, but their histograms are

the same. Suppose that each image is blurred with a 3 × 3 averaging mask.

Explain whether the histograms of the blurred images would be equal. If your

answer is no, sketch the two histograms.

Fig. for Q. 8(b)

(15)

L-4/T-2/EEE Date: 12/01/2021



Sub: EEE 451 (Processing and Fabrication Technology)




There are 2 pages in this question paper.

SECTION – A


[All the symbols have their usual meanings. Assume reasonable values for missing data.]

1. (a) Briefly discuss the various technologies used to fabricate plasmonic devices.

What is a HEPA filter and how does it help in a cleanroom?

(10+5)

(b) How does chemical mechanical polishing (CMP) help improve wafer

quality? What are the differences between CZ and Float Zone methods of wafer

fabrication?

(8+7)

2. (a) Draw the process sequence diagram (sequential processes) involved in

“Phosphoric Oxide Deposition and Cap Oxidation”.

(12)

(b) Explain vapor phase growth steps in terms of gas phase decomposition,

adsorption, surface diffusion.

(18)

3. (a) Describe ion implantation technique with a proper diagram. (20)

(b) Describe the operation of Molecular Beam Epitaxy.

(10)

4. (a) Why is plasma etching preferred in some applications over wet etching?

How does LPCVD differ from APCVD?

(5+5)

(b) Explain the Deal-Grove model of oxidation in detail. (20)

= 2 =

SECTION – B


[All the symbols have their usual meanings. Assume reasonable values for missing data.]

5. (a) Why silicides are used for ohmic contacts? Compare them with copper for

metal contacts.

(15)

(b) Discuss the functions of IC packaging. (15)

6. (a) Discuss potential hazards from cleanroom pollutants. (15)

(b) Discuss the necessity of hard bake during ten-step process.

(15)

7. (a) What are the two types of photoresists? Discuss their differences and

chemical compositions.

(15)

(b) What are the aligner selection criteria? Discuss the differences between

contact aligner and proximity aligner.

(6+9)

8. Write short notes on - (30)

(a) Dry oxidation

(b) Extreme Ultraviolet Lithography

(c) Two color pyrometry

L-4/T-2/EEE Date: 09/01/2021


L-4/T-2 B.Sc. Engineering Examination 2018-19

Sub: EEE 457 (VLSI Circuits and Design II)


The figures in the margin indicate full marks


There are 4 page(s) in this question paper

SECTION – A


All the symbols have their usual meaning

Assume reasonable values for missing data

1. (a) Discuss the difference between device scaling and interconnect scaling in

terms of delay parameters. In a 1 um CMOS process, Wn/Ln=10 um/1 um,

Wp/Lp=20 um/1 um, tgox=30 nm, εox=0.35 pF/cm, Cint=2 pF/cm. Calculate the

interconnect length at which interconnect capacitance becomes comparable

with gate capacitance.

(20)

(b) Is the current density of devices affected by CMOS scaling trends? Explain

your answer using the Dennard principle.

(10)

2. (a) Design a four-input NOR gate with transistor widths chosen to achieve

effective rise and fall resistance equal to that of a unit inverter. Then sketch

equivalent circuits for falling and rising output transitions and estimate the

gate’s worst-case propagation delay if the output is loaded with h identical

gates like itself. Use appropriate delay models.

(20)

(b) Explain how high-k dielectric can assist the progression of CMOS

technology.

(10)

3. (a) Using basic delay models, estimate the frequency of a 15-stage ring

oscillator. Assume it uses a 0.6 um process with τ = 60 ps.

(12)

= 2 =

(b) Prove that for the minimum-size repeaters to be useful for improving

wire delay, the RC constant of the wire must be at least seven times the delay of

a minimum-size buffer.

(18)

4 (a) Explain the relative merits and limitations of cascode voltage switch logic

over other circuit families.

(10)

(b) Write short notes on the following topics related to a CMOS process flow:

ILD-1, STI, Tungsten Plug.

(20)

SECTION – B


All the symbols have their usual meaning

Assume reasonable values for missing data

5. (a) A 32 bit adder have to be designed using carry select approach. Delay of the

one bit adder circuit is 2 nsec and the delay of the multiplexer is 1 nsec.

Determine the optimum block size and number of adders in each block. Draw the

schematic diagram of the second block.

(15)

(b) Sketch the partial products used by a radix-4 Booth-encoded multiplier to

compute (18)10 (12)10 and show the generation of the corresponding final

products. Design a Booth encoder using Xi, X2i, and Mi where Xi is true for Y,

X2i is true for 2Y and Mi is true for negative partial products.

(15)

6. The figure below shows the self-bypass paths for six separate ALUs of a

microprocessor. The path for one of the ALUs begins at registers containing the

inputs to an adder, as shown in the Fig. for Ques. 6. The adder must compute the

sum. A result multiplexer chooses between this sum, the output of the logic unit,

and the output of the shifter. Then a series of bypass multiplexers selects the

inputs to the ALU for the next cycle. The early bypass multiplexer chooses

among results of ALUs from previous cycles and is not on the critical path. The

8:1 middle bypass multiplexer chooses a result from any of the six ALUs, the

early bypass mux, or the register file. The 4:1 late bypass multiplexer chooses a

result from either of two results returning from the data cache (Dc1. Dc2), the

middle bypass mux result (Bypass), or the immediate operand (Imm) specified

by the next instruction. The late bypass mux output is driven back to the ALU to

use on the next cycle. Because the six ALUs and the bypass multiplexers occupy

a significant amount of area, the critical path also involves 2 mm wires from the

result mux to middle bypass mux and from the middle bypass mux back to the

late bypass mux. The propagation delays and contamination delays of the path

are given in Table for Ques. 6.

= 3 =

Fig. for Ques. 6

Table for Ques. 6

Element Propagation delay Contamination delay

Adder 450 ps 100 ps

Result Mux 50 ps 35 ps

Early Bypass Mux 110 ps 95 ps

Middle Bypass Mux 80 ps 55 ps

Late By pass Mux 70 ps 45 ps

2mm wire 100 ps 65 ps

(A) (i) Define setup time and hold time of flip-flops and propagation delay and

contamination delay of logic circuits.

(ii) Suppose the registers are built from flip-flops with a setup time of 62 ps,

hold time of –10 ps, propagation delay of 90 ps, and contamination delay of

75 ps. Calculate the minimum cycle time Tc at which the ALU self-bypass

path will operate correctly.

(B)(i) If the earliest input to the Late bypass multiplexer is the imm value coming

from another flip-flop, will this path experience any hold time failure?

(ii) If the ALU self-bypass path uses pulsed latches in place of flip-flops, will

it have any hold time problems? How can the hold time problem, if it exists,

be eliminated? The pulsed latch has a pulse width of 150 ps, a setup time of

40 ps, a hold time of 5 ps, a clock to Q propagation delay of 82 ps and a clock

to Q contamination delay of 52 ps, and a D-to-Q propagation delay of 92 ps.

(6)

(6)

(6)

(12)

7 (a) A NAND based decoder circuit is used as row decoder of a 4K memory block.

Calculate number of transistors required if CMOS circuit is used with (i) no pre-

decoding and (ii) 2-bit pre-decoding.

(15)

(b) Draw the transistor level circuit diagram of a 4x4 bit NOR ROM which stores

the following data:

(15)

= 4 =

0 1 1 1

1 0 0 1

0 1 0 0

0 1 1 1

8 (a) Consider the circuit in Fig. for Ques. 8. Determine the activity factors at

each node in the circuit (node n1, n2 and y) assuming the input

probabilities PA = PB = PC = PD = 0.5. Also calculate the dynamic power

dissipation of each gate G1, G2 and G3. Given, output capacitance at

node n1 and n2 is 10 fF each and 15 fF at node y. Given : power Supply

VDD= 5V, the operating frequency f = 10 MHz.

A

C

D

B

n1

n2

y

G1

G2

G3

Fig. for Ques. 8

(20)

(b) Suppose that you are asked by your manager to reduce the power dissipation

of the circuit shown in Fig. for Ques. 8 as much as you can. Describe what

possible steps you can take to reduce the power dissipation. Assume that the

process has thin oxide and thick oxide transistors and each of these transistors

has two different types of Vt (high-Vt and low-Vt).

(10)

First Row

First Column

L-4/T-2/EEE Date: 16/01/2021



Sub: EEE 459 (Optoelectronics)





SECTION – A




1. Draw and discuss photoluminescence spectra and absorption coefficient near and

above the fundamental absorption edge for GaAs1-xPx semiconductor. Draw

schematic E-k diagram and estimate band gap of the semiconductor. Comment

on the layer grown on GaAs substrate and compare with the layer grown on GaP

substrate in terms of optical application.

Given: x= 0.2+ Last 3 digits of your Student ID No.× 10-3

(30)

2. Design a tandem three-junction solar cell using ternary and quaternary

semiconductors. Name the semiconductors used and mention their band gaps.

Justify to your selection of semiconductors in terms of absorption properties and

lattice matching in different layers.

Estimate efficiency of the proposed tandem solar cell mentioning various loss

factors.

(30)

3. (a) Design a Laser Diode to operate at X nm. Name the materials used and make justification regarding the materials chosen

at different layers.

Discuss the merits and demerits of the designed LD.

Given:

X =500 + 2Y nm (Y= Last 3 digits of your Student ID No.).

(20)

= 2 =

(b) Consider a Fabry-Perot laser diode operating at W nm. The threshold current

is 6 mA. At I = 25 mA, the output optical power is 6 mW and the voltage across

the diode is 1.2 V. Calculate

(i) external quantum efficiency (QE),

(ii) power conversion efficiency, and

(iii) slope efficiency of the diode.

What is the forward diode current that gives an output optical power of 3 mW?

W= 400 + 5Y nm (Y= Last 3 digits of your Student ID No.).

(10)

4 (a) Design a lithium niobate (LiNbO3) Pockels cell phase modulator, that will

operate at a free-space wavelength of 1.1+X μm and will provide a phase shift

ΔΦ of ℼ (half wavelength) between the two field components propagating

through the crystal for an applied voltage of 12 V?

Given:

At λ = 1.1+X μm, LiNbO3 has no ≈ 2.21, r22 ≈ 5 × 10-12 m V-1.

X= Last 2 digits of your Student ID No.× 10-2

(20)

(b) Consider a ZnTe crystal which rotates the optical field of the 633 nm

polarized laser beam from a He-Ne laser. Calculate the necessary magnetic field

for a rotation of Y° over a crystal length 20 mm.

Given:

Y= 2+Last 2 digits of your Student ID No.× 10-2

ZnTe have Verdet constant of about 188 rad T-1 m-1, for light having wavelength

633 nm.

(10)

SECTION – B




5. (a) How the drawbacks of homojunction LED are overcome in double-

heretostructure LED? What should be the optical properties of typical materials

used in confining layers of surface emitting LEDs? Discuss the function of these

confining layers.

(12)

= 3 =

(b) The ternary alloy In1-xGaxAsyP1-y grown on an InP crystal substrate is a

suitable semiconductor material for infrared wavelength LED. The device

requires that the InGaAsP layer is lattice matched to InP crystal substate to avoid

crystal defects in the InGaAsP layer. This in turn requires that y ≈ 2.2x. The

bandgap Eg for this ternary alloy in eV at 298 K is given by the empirical

relationship,

Eg ≈ 1.35 - 0.72y + 0.12y2; 0 ≤ x ≤ 0.47

(i) Calculate the compositions of InGaAsP ternary alloy for peak emission

at a wavelength of 1.55 µm.

(ii) For 1.55 µm peak wavelength, what is the wavelength difference between

the half intensity points in the output spectrum? Assume that the spread

in the photon energies Δ(hν) ≈ 3kBT between the half intensity points.

(18)

6. (a) Explain how recombination center is produced by nitrogen incorporation in

GaAs1-xPx. Discuss the effect of nitrogen incorporation in GaAs1-xPx LED in

terms of emission wavelength and quantum efficiency.

(14)

(b) Calculate (i) radiative recombination efficiency, (ii) external quantum

efficiency, and (iii) extraction efficiency for a GaAs n+-p junction 867 nm

infrared LED. Consider, the active region i.e., p-side of the junction have doping

concentration of 5×1016 cm-3 and the nonradiative lifetime is about 50 ns. When

1.2 V voltage was applied across the LED, a forward current was measured to be

30 mA, and the emitted optical power was found to be 6 mW. Assume, Br = 7 ×

10-10 cm2s-1 for GaAs.

(16)

7. (a) Suggest a structure with necessary schematics for overcoming the total

internal reflection problem in the LEDs. Among the homojunction,

heterojunction and quantum well LEDs, which one is expected to have the

narrowest emission linewidth?

(15)

= 4 =

(b) If semiconductor slab of length l, width w, and depth d shown in Fig for Q.

7(b) acts as a photoconductive detector, mention the factors on which the

photocurrent and photoconductive gain of this detector depend? Which factor

limits the speed of response of the detector?

Fig for Q. 7(b)

(15)

8. (a) With necessary schematics and band diagrams briefly discuss the operational

principle of a Schottky junction photodiode. Schottky junction photodiodes are

better suited for detecting short-wavelength light than pn and pin photodiodes –

justify the statement. Why Schottky junction photodiodes can be significantly

faster than pn and pin photodiodes?

(15)

(b) A Ge APD has a quantum efficiency of 50% at 1.55 µm in the absence of

multiplication. It is biased to operate with a multiplication of 20. (i) Calculate the

photocurrent if the incident optical power is 50 nW. (ii) What is the responsivity

when the multiplication is 20?

(15)

L-4/T-2/EEE Date: 23/01/2021



Sub: EEE 461 (Semiconductor and Nano Device)





SECTION – A




1. (a) What is collisional broadening of electronic states? Comment on relative collisional broadening of electronic states in one-, two-, and three-dimensional materials.

(12)

(b) Derive the electron transmission coefficient in a resonant tunneling diode. How will the transmission coefficient be affected if an electric field is applied across the resonant tunneling diode?

(18)

2. (a) How the perturbation changes the Hamiltonian for an electronic state due to the interaction with photons? How is it different from that when the interaction happens due to phonons?

(12)

(b) An electron is in the second state of a GaAs quantum well of width 15 nm, which can be treated as an infinitely deep one-dimensional system. Suddenly the middle 5 nm becomes 100 meV deeper. Use the golden rule to derive an expression for the rate at which the electron scatters into the first, third and fourth states, and evaluate the rate for the lowest allowed transition.

(18)

3. (a) Discuss the dependence of phonon scattering rate with temperature. Comment on the comparison between interband transitions due to phonons vs. intersubband transitions due to phonons.

(12)

(b) Calculate the transmission and reflection flux coefficient for an electron of energy E, moving from left to right, impinging normal to the plane of a semiconductor heterojunction potential barrier of energy V0, where the effective electron mass on the left-hand side is m1 and the effective electron mass on the right-hand side is m2.

(18)

= 2 =

If the potential barrier energy is V0 = 1.5 eV and the ratio of effective electron mass on either side of the heterointerface is m1/m2 = 3, at what particle energy is the transmission flux coefficient unity?

4 (a) What determines the selection rules for optical transitions at frequency w between states i and j?

(10)

(b) An electron is initially in the ground state of a one-dimensional rectangular potential well for which V(x) = 0 in the range 0 < x < L and V(x) = ¥ elsewhere. The ground state energy is E1 and the first excited state energy is E2. At time t = 0, the system is subject to a perturbation 𝑉!𝑥"𝑒#$/&. Calculate and plot the probability of finding the particle in the first excited state as a function of time for 𝑡 ≥ 0. In your plot, normalize time to units of t.

(20)

SECTION – B



5. (a) The dispersion relation of a linear diatomic lattice is given by the following equation where different symbols have their usual meanings.

Considering M > m, derive the dispersion relation for the acoustic branch.

(18)

(b) According to Kronig-Penny model, do you expect the energy gap to increase or decrease as electron energy increases? Why? Draw the reduced zone representation of the E-k diagram in accordance with your answer.

(12)

6. (a) Consider the following expression of internal energy where different symbols have their usual meanings.

Based on this relation, derive the expression of specific heat and discuss how your derivation deviates from experimental observation.

(18)

(b) ‘Thermal blurring of electron distribution is more significant in the degenerate limit’- do you agree with this statement? Justify your answer with necessary equations and diagram.

(12)

7. (a) Consider a finite potential well which has a single energy state E1 and corresponding wavefunction Ψ1 under equilibrium. Suppose a small external DC voltage Vo is applied to it. Based on perturbation theory, derive the expression of second-order corrected energy state and first order corrected wavefunction.

(20)

= 3 =

(b) Define Reststrahlen peak. Between NaCl and GaAs, in which material do you expect Reststrahlen effect to be stronger and why?

(10)

8. (a) Based on tight-binding model, show with necessary equations how bringing a pair of wells of identical atoms causes the energy level to split into two levels.

(20)

(b) For a 1-D monoatomic lattice, draw the dispersion relation where the two transverse modes correspond to two different phase velocities. For what kind of crystals phase velocities of the two transverse modes will be identical?

(10)

L-4/T-2/EEE Date: 16/01/2021



Sub: EEE 483 (High Voltage Engineering)





SECTION – A




1. (a) In high voltage (HV) AC generation the deviation of HV from sinusoid of

fundamental frequency should be limited to 5% of peak value. What problem

will a higher deviation cause and how?

(7)

(b) In what type of high voltage testing would you consider using a resonance

transformer? In such tests the test object is considered pure capacitive. What is

the basis of such consideration? What effect does this capacitance have on the

test?

(c) Briefly present the operating characteristics of the resonant transformer

circuit with variable test frequency.

(8)

(15)

2. (a) In electrical breakdown of gases, it is noted that the effectiveness of

ionization by electron impact depends upon energy that an electron can gain

along mean free path in the direction of field.

(i) Explain “mean free path.” What effect does temperature and pressure

have on it? Comment on the mean free path of an atom in its own gas and

that of an electron in a gas.

(ii) What is collision cross-section? What is its relation with mean free path?

What factors may affect collision cross-section?

(15)

(b) Townsend introduced two ionization coefficients to explain electrical (15)

= 2 =

breakdown of gases. Discuss the physical phenomena these coefficients signify.

3. Explain the „Kanal‟ mechanism of spark.

(30)

4. (a) Present a qualitative comparison of the characteristics of anode corona and

cathode corona in atmospheric air.

(20)

(b) For purpose of coordinating electrical stresses with electrical strengths two

characteristics are important, overvoltage distribution and insulation breakdown

probability. How are these used to assess insulation failure risk?

(10)

SECTION – B


5. (a) With neat diagrams, describe the operating principle of three different

voltage doubler circuits. For each circuit, mention when the diodes conduct and

explain how voltage is doubled at the output terminal.

(20)

(b) Derive the expression of Ripple Factor for Cockcroft–Walton type voltage

doubler circuit.

(10)

6. (a) With necessary diagrams explain how peak value of ac voltage and impulse

voltage can be measured.

(20)

(b) Mention the principal sources of error in Chubb–Fortescue method.

(10)

7. (a) Explain how breakdown occurs in solid dielectrics due to edge effect,

treeing and tracking. How can these be prevented?

(b) Discuss the effect of the frequency of the applied voltage on thermal

breakdown strength of a solid dielectric.

(25)

(05)

8 “Presence of solid and gaseous impurities in liquid insulation has a profound

effect on breakdown strength of liquids”- Briefly explain the mechanism of

liquid breakdown in presence of such impurities.

(30)

L-4/T-II/EEE Date: 12/1/2021



Sub: EEE 487(Nuclear Power Engineering)





SECTION – A


All the symbols and abbreviations have their usual meanings.

Assume reasonable values for missing data (if any).

1. (a) What do you mean by enrichment? What are the methods for enrichment

based on atomic mass number of isotopes?

(12)

(b) Derive a simple formula relating feed, product and waste (tail) per day in a

uranium enrichment plant.

(c) How many kg of natural U-238 is required in an enrichment process to

obtain 1 kg product with 5% U-235? Given that the abundance of U-235 in

natural uranium is 0.7% and U-235 must not be more than 0.3% in the waste.

(10)

(8)

2. (a) What is meant by off-site and on-site power systems in relation to a nuclear

power plant? What are their functions?

(12)

(b) What is Class 1E subsystem? What are the main functions of the I&C

system in a nuclear power plant?

(c) Explain the terms LOOP, SBO and LOCA.

(12)

(6)

3. (a) Explain the impacts of grid frequency and voltage on a nuclear power plant

operation.

(15)

(b) Explain the impacts of the size of a nuclear power plant unit on the grid

operation.

(15)

=2 =

4 (a) What are the situations in which a nuclear power plant has to “Scram”? (5)

(b) Explain the root causes of Three Mile Island, Chernobyl and Fukushima

Daiichi nuclear power plant accidents.

(c) Explain nuclear fuel cycle, front end, back end, open cycle and closed cycle.

(15)

(10)

SECTION – B


All the symbols and abbreviations have their usual meanings.

Assume reasonable values for missing data (if any).

5. (a) What is nuclear fission and fusion? With necessary equations and diagrams,

explain the fission process in U-235 nuclei.

(15)

(b) Assume 80% of neutrons are absorbed by U-235 cause fission and rest

being absorbed by the non-fission capture to produce an isotope U-236. Each

fission of U-235 yields 190 MeV of useful energy. Estimate the fuel

consumption of U-235 per hour to produce 100 MW of thermal power.

(15)

6. (a) With necessary diagrams, explain the steam generation and electrical power

production in a nuclear power plant.

(15)

(b) (i) What is waste heat rejection in a nuclear power plant?

(ii) A nuclear reactor produces 3000 MW of thermal power. The water enters at

300 oC and leaves at 325 oC. Assuming the water is at 2000 psi and 600oF

(PWR conditions), find the amount of circulating water in litres/minute to cool

the reactor. Given that the specific heat is 6.06 x 103 J/(kg. oC) and the density

is 687 kg/m3.

(15)

7. (a) Write down the attributes of Generation III, Generation III+ and Generation

IV nuclear reactors.

(15)

(b) With necessary illustrations, briefly explain the construction and operation

of a Pressurized Water Reactor (PWR).

(15)

=3 =

8. (a) A 1000 MWe nuclear unit built for $2 billion achieves a capacity factor of

90% with a burn-up of 30,000 MWD/MTU. Fuel costs are $1025/kg and O&M

expenses are $60 million/year. Find the electricity generation cost of this plant

by considering a levelized fixed charge rate of 17%/year.

(10)

(b) Write down the key characteristics of six next generation nuclear reactors.

Explain the role of next generation reactors in hydrogen economy.

(20)

L-4/T-II/EEE Date: 23/1/2021



Sub: EEE 489(Smart Grid)





SECTION – A


All the symbols and abbreviations have their usual meanings

1. (a) Define the grid reliability indices: SAIFI, SAIDI and ASAI. What are the

IEEE standard limits for SAIFI and SAIDI? How can smart grid help avoid

exceeding these limits?

(18)

(b) Explain transactive energy (TE). What are the differences between TE and

DR (demand response)?

(12)

2. (a) Define grid resiliency and grid hardening. How can smart grid help achieve

these?

(14)

(b) What is the difference between “consumer” and “prosumer”? Explain how

a photovoltaic (PV) plant can be controlled to reduce the output at a given

irradiance, in response to a command from the load dispatch centre ?

(16)

3. (a) What are the bay controllers in a substation? Using diagrams show two

types of communication between bay controllers and the substation computer or

RTU.

(14)

(b) Using a diagram show a typical architecture for communication between a

smart meter and the utility server.

(c) What type of communication technology is employed between a DCU and a

smart meter and between a DCU and the gateway?

(10)

(6)

=2 =

4 (a) There are 250,000 smart meters in a power distribution system and there are

1000 concentrators. The data network is designed such that if one link from a

concentrator fails, another link will transmit the data. Each smart meter sends 3

bytes of data every second. Then what should be the average data rate through a

data link between the concentrator and the server?

(10)

(b) Write down the application areas of following communication technologies

in smart grid.

Bluetooth, ZigBee, 6LoWPAN, WiMax, FOC.

(c) What is protocol? Mention the names of widely used protocols along with

their applications in power system or smart grid communication.

(10)

(10)

SECTION – B


All the symbols and abbreviations have their usual meanings

5. (a) What is a smart grid? What are the motives behind implementing a smart

grid? Compare a traditional grid with a smart grid in terms of power generation,

transmission, distribution and control.

(15)

(b) Provide an overview of the following technological requirements for a smart

grid.

(i) Information and communication technologies, (ii) Sensing, measurement,

control and automation technologies and (iii) Power electronics and energy

storage systems.

(15)

6. (a) Present a list of electrical and non-electrical energy storage systems.

Describe the possible solutions to uncertainties of power output from variable

renewable energy sources in a power system.

(15)

(b) What are the differences between Home-Area Network (HAN) and Internet

of Things (IoT)? Provide a brief description of a HAN.

(15)

=3 =

7. (a) With necessary diagrams, explain the following services provided by

Demand-Side Integration (DSI).

(i) Load shifting, (ii) Valley filling, (iii) Peak clipping and (iv) Energy

efficiency improvement.

(15)

(b) Define price elasticity of demand and elasticity of substitution. How can

DSI enable frequency control support in a smart grid?

(15)

8. (a) What is a microgrid? Define self-healing and restoration process. (10)

(b) Explain (i) the hierarchical control and (ii) islanded mode operation of any

typical microgrid.

(20)

L-4/T-2/EEE Date: 16/01/2021



Sub: EEE 497 (Telecommunication Networks)


The Figures in the right margin indicate full marks



SECTION – A



1. (a) Apply the Dijkstra algorithm and determine the shortest paths from the

source node A to the destination nodes E and G for the network shown in Fig.

1(a).

Fig. 1(a)

(15)

(b) Suppose, BUET is granted a block of IP addresses with the beginning

address 172.196.120.0/22. The number of hosts in some of departments of

BUET is given in Table 1. Design the sub-networks for the mentioned

departments of BUET providing sub-network address, sub-net mask and sub-

network host addresses.

(15)

B

A

C

E

F

D

G

8

2

2

2

2

1

5

11

1

3

6

5

= 2 =

Table 1: No. of hosts for each Department

Dept. Name No. of Hosts

ME 30

CE 56

CSE 130

EEE 100

BME 12

.

2. (a) In a TCP operation, the initial values of smoothed round-trip time and round

trip time variation are 100 ms and 10 ms, respectively. The acknowledgement

times of a sequence of frames are 100 ms, 90 ms, 80 ms, 100 ms, 70 ms, 90 ms,

80 ms, 60 ms and 90 ms. Calculate the retransmission time out (RTO) at the

end of all the sequence. Assume, α=7/8 and β=3/4.

(15)

(b) In a TCP Tahoe operation, the initial value of slow start threshold is 64 kB

and the length of each frame is 2 kB. The buffer size is assumed to be 80 kB

where the frame loss starts. Draw the congestion window versus transmissions

rounds (RTTs) showing slow start, additive increase and multiplicative

decrease phases.

(15)

3. (a) How do the RTP and RTCP work in VoIP application along with UDP?

Explain, why TCP does not perform well for VoIP.

(15)

(b) Consider an H.323 client X wants to perform VoIP communication with an

H.323 client Y. Explain, the operation of connection set up, communication and

connection release between the VoIP clients.

(15)

4 (a) Consider the two ISDN Layer-3 entities A and B. Describe how the

association is completed by the two entities.

(15)

(b) Explain the meaning of the following 5 byte ATM header.

1010 1001

1001 0101

1010 1010

1101 0110

1010 0101

(15)

= 3 =

SECTION – B



5. (a) What could be the possible problems if packet switching is used for voice

application and circuit switching is used for data application? Explain with

justification.

(10)

(b) Networks can be classified as access networks and backbone networks. With

justifications and mentioning the advantages and disadvantages, explain which

topologies should be used and should not be used for access networks and

backbone networks.

(20)

6. (a) The multistage switching system shown in Fig. for Q. No. 6(a) has 12 inlets.

Determine -

(i) the total number of cross-points used in the system.

(ii) the blocking probability using Lee graph when the utilization of an

inlet is 0.2.

Fig. for Q. No. 6(a)

(20)

(b) Explain how the time switching is operated in digital TDS. (10)

7. (a) In a slotted ALOHA system, the packet arrival rate is λ packets/sec. The

length of each packet is 1000 byte and the data transmission rate is 2 Mbps when

a packet is transmitted. Determine the value of λ for achieving the maximum

throughput. Also, determine the maximum throughput in packets/sec and Mbps.

(10)

(b) Explain how the CSMA/CA MAC protocol manages the effect of increasing

and decreasing the numbers of users in a WLAN. Also explain the trade-off

between the hidden terminal problem and exposed terminal problem in a CSMA

system with carrier sensing range.

(20)

= 4 =

8. (a) Distinguish among SDH, WDM and OTN in terms of operating principle, data

rate and applications.

(15)

(b) Explain how the routing and forwarding of MPLS technology are different

from the routing and forwarding of IP technologies.

(15)

l-4/t-2/eee date: 09/01/2021 - lib.buet.ac.bd:8080

Documents